Pharmacovigilance and expedited drug approvals

Pharmacovigilance in High-Income Countries: Current Developments and a Review of Literature

The world bank has classified 80 economies based on their Gross National Income (GNI) per capita as High-Income. European Medicines Agency (EMA), Food and Drug Administration (FDA), and Pharmaceuticals and Medical Devices Agency (PMDA) are the major regulatory stakeholders driving global pharmacovigilance regulations. The purpose of this article is to describe pharmacovigilance systems and processes in high-income countries, particularly those that are also members of the International Conference on Harmonization (ICH). All high-income countries are members of the WHO PIDM. The income level of a country has a direct relationship with medicine safety measures. All ten pioneering members of the Uppsala monitoring centre are from high-income countries and were the first responders after the thalidomide tragedy by making drug evaluation committees, introducing the ADR reporting forms and taking safety measures. Despite access to the VigiBase, some countries have separate databases for managing and analyzing data like Canada Vigilance online database, FDA Adverse Event Reporting System, the French pharmacovigilance database and European Union's system Eudravigilance. All high-income countries have robust pharmacovigilance systems. USFDA and EMA are the world leaders in the field of pharmacovigilance. Most high-income countries follow EMA guidelines. Medicine safety is directly influenced by a country's income level.

Adverse drug reactions

An adverse drug reaction (ADR) is defined as a response to a medicinal product which is noxious and unintended. ADRs are an important cause of morbidity and mortality and increase health costs. The pharmacovigilance systems allow the identification and prevention of the risks associated with use of a drug, especially of recently marketed drugs; they detect signals from data of the global ADR register and also support decisions taken by regulatory agencies in different countries. Only a few drugs are withdrawn from the market, mainly due to hepatotoxicity. Spontaneous notification of ADR is the cheapest, simplest and most used method to recognize new safety drug problems, under-reporting being its main limitation. The future of pharmacovigilance and ADRs will include a higher involvement of patients, doctors, health authorities and pharmaceutical companies, and the use of new technologies.

A Machine-Learning Algorithm to Optimise Automated Adverse Drug Reaction Detection from Clinical Coding

INTRODUCTION

Adverse drug reaction (ADR) detection in hospitals is heavily reliant on spontaneous reporting by clinical staff, with studies in the literature pointing to high rates of underreporting [1]. International Classification of Diseases, 10th Revision (ICD-10) codes have been used in epidemiological studies of ADRs and offer the potential for automated ADR detection systems.

OBJECTIVE

The aim of this study was to develop an automated ADR detection system based on ICD-10 codes, using machine-learning algorithms to improve accuracy and efficiency.

METHODS

For a 12-month period from December 2016 to November 2017, every inpatient episode receiving an ICD-10 code in the range Y40.0-Y59.9 (ADR code) was flagged for review as a potential ADR. Each flagged admission was assessed by an expert pharmacist and, if needed, reviewed at regular ADR committee meetings. For each report, a determination was made about ADR probability and severity. The dataset was randomly split into training and test sets. A machine-learning model using the random forest algorithm was developed on the training set to discriminate between true and false ADR reports. The model was then applied to the test set to assess accuracy using the area under the receiver operating characteristic (AUC).

RESULTS

In the study period, 2917 Y40.0-Y59.9 codes were applied to admissions, resulting in 245 ADR reports after review. These 245 reports accounted for 44.5% of all ADR reporting in our hospital in the study period. A random forest model built on the training set was able to discriminate between true and false reports on the test set with an AUC of 0.803.

CONCLUSIONS

Automated ADR detection using ICD-10 coding significantly improved ADR detection in the study period, with improved discrimination between true and false reports by applying a machine-learning model.